In this paper, we investigate data transmission schemes from sources to sinks through multiple-relay networks. We propose myopic coding schemes for data relaying, whereby each node only needs to consider a few neighboring nodes during encoding and decoding. These schemes are more robust to topology changes than existing omniscient coding schemes, in which all nodes are considered in the coding design at each node. When myopic coding is deployed, the encoding and decoding at a node will not be affected when there are changes in remote nodes. We compute achievable rate regions for these myopic coding schemes and compare them with that of the omniscient schemes. Using Gaussian channels as examples, we show that, the two-hop myopic scheme performs as well as the omniscient scheme in a five-node multiple-relay channel and close to the omniscient scheme in a six-node multiple-relay channel, when nodes transmit at low power. Myopic schemes also mitigate the high computational complexity and large buffer/memory requirements of omniscient coding schemes.